I need to repeat a certain test a number of times $n$ and then output the average of the values obtained in these $n$ tests. I have been told that I should use ANOVA to determine $n$ such that the obtained mean value is a significant one. What are the usual ways in which one decides how many times to repeat a test?
A bit more context, although not necessarily crucial to my answer:
I am currently developing a genetic algorithm for the TSP problem. I need to show how the algorithm performs with different values of its parameters (population size, instance size, recombination probabilities, etc.). I have determined which are the tests I want to perform , but I don't know how many times I should repeat them. In other words, for a certain value of the parameters, I will repeat the test $n$ times and then show what is the average optimal solution given by the algorithm. My question is, how can I choose $n$ so that the resulting mean is significative?